README for Replication Folder

Manuscript Title: Partisan and Ideological Bias Among the Attentive Public: Evidence from Witness Slips in the Illinois General Assembly

Authors: Michael Kistner and Michael Pomirchy

Date: December 17, 2025

---

Computing Environment

R version 4.3.1 (2023-06-16 ucrt) -- "Beagle Scouts"
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows 11 x64 (build 22631)

R Packages Used:
- tidyverse 2.0.0
- ggplot2 3.5.2
- dplyr 1.1.4
- tidyr 1.3.1
- readr 2.1.4
- stringr 1.5.1
- purrr 1.1.0
- forcats 1.0.0
- lubridate 1.9.4
- cowplot 1.2.0
- fixest 0.11.1
- janitor 2.2.1
- scales 1.4.0

---

Data Files

1. "Slip Data (Original Download).csv" - CSV format
   - ORIGINAL DATA SOURCE(S): Scraped from the Illinois General Assembly website (https://my.ilga.gov)
   - Contains 2,266,522 observations across 7 variables
   - Variables include: bill identifier, witness name, employer, interest group affiliation, hearing date and committee, position taken (support/oppose/no position), and amendment information
   - Raw witness slip data as originally downloaded from the Illinois General Assembly website covering legislative sessions 2013-2022

2. "Bill Dataframe.Rda" - R data format
   - ORIGINAL DATA SOURCE(S): Constructed from cleaning Slip_Data_Original_Download.csv
   - Contains 49,779 observations across 11 variables
   - Variables include: bill identifier, description, session, chamber, sponsor name and party, amendment status, and counts of witness slips by position (support/oppose/no position)
   - Bill-level dataset aggregating witness slip activity for each piece of legislation in the Illinois General Assembly

3. "Ideal Points Dataframe.Rda" - R data format
   - ORIGINAL DATA SOURCE(S): Ideal points estimated using Bayesian IRT model; legislator ideal points anchored to Shor-McCarty scores (https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/GZJOT3)
   - Contains 12,562 observations across 12 variables
   - Variables include: witness/legislator ID, estimated ideal point (D1), party affiliation, Shor-McCarty score (for legislators), actor type, witness name, employer, standardized group affiliation, number of slips filed, and number of votes used for scaling
   - Individual-level dataset containing ideology estimates for witnesses and legislators placed on the same liberal-conservative dimension

4. "Witness Dataframe.Rda" - R data format
   - ORIGINAL DATA SOURCE(S): Constructed by aggregating witness-level information from Slip_Data_Original_Download.csv
   - Contains 517,193 observations across 10 variables
   - Variables include: unique witness ID, full name, first name, last name, total number of slips filed, standardized group affiliation, binary group affiliation indicator, and partisan lean measures (net support for Republican bills, Democratic bills, and overall partisan lean)
   - Witness-level dataset containing aggregated information about each individual who filed witness slips

5. "Witness Slips Dataframe Cleaned.Rda" - R data format
   - ORIGINAL DATA SOURCE(S): Cleaned version of Slip_Data_Original_Download.csv with standardized variables and additional bill/sponsor information from ILGA website
   - Contains 2,155,272 observations across 13 variables
   - Variables include: session, full bill name, committee, hearing date, sponsor party, amendment name, amendment status, numeric position indicator, witness ID, witness name, number of slips filed by witness, binary group affiliation indicator, and standardized group name
   - Slip-level dataset with each row representing one witness slip filed on a bill or amendment

---

Code Files

1. Replication_Code.R
   - Reproduces Figures 1-7 from the manuscript
   - Loads all five main datasets (Bill_Dataframe, Ideal_Points_Dataframe, Slip_Data_Original_Download, Witness_Slips_Dataframe_Cleaned, Witness_Dataframe) plus Shor-McCarty ideal points
   - Figure 1: Distribution of witness slips per bill (log scale histogram) and distribution of positions taken (bar chart showing support, oppose, and no position)
   - Figure 2: Partisan direction of witness slip support at bill-level (net support for Republican vs. Democratic bills) and witness-level (average partisan lean across all witnesses)
   - Figure 3: Ideology of witnesses and slips relative to state legislators, showing mean witness and mean slip ideal points compared to median Democratic and Republican legislators
   - Figure 4: Ideological distributions comparing Illinois state legislators (Democrats and Republicans) to witnesses using density curves
   - Figure 5: Differences between group-affiliated and non-affiliated witnesses across four panels (bill-level partisanship, witness and slip ideology, witness-level partisanship, and distribution of witness ideology)
   - Figure 6: Scatterplot showing witness slip involvement and ideology by interest group, with circle size indicating number of unique witnesses and y-axis showing total slips filed
   - Figure 7: Distribution of witness ideology within groups using dot plots for all groups with 25+ scaled witnesses, showing ideological heterogeneity within organizations

---